AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer

Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, Peter J. Liu

JMLR 2020 · Large language models › Language, safety and interpretability · approach: Text generation and evaluation metrics · 13,452 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add unsupervised pre-training (ablation)T5 baseline · GLUEaverage score: 66.22 -> 83.28Improved
add unsupervised pre-training (ablation)T5 baseline · SQuADexact match: 50.31 -> 80.88Improved
add unsupervised pre-training (ablation)T5 baseline · WMT En-FrBLEU: 39.77 -> 39.82No clear effect
use encoder-decoder instead of decoder-only language model (denoising) (ablation)T5 · GLUEaverage score: 74.70 -> 83.28Improved
use denoising instead of LM objective (ablation)encoder-decoder · GLUEaverage score: 79.56 -> 83.28Improved
share parameters across encoder and decoder (ablation)encoder-decoder · GLUEaverage score: 83.28 -> 82.81No clear effect
use 12 layers instead of 6 per stack (ablation)encoder-decoder · GLUEaverage score: 80.88 -> 83.28Improved
use BERT-style denoising instead of deshuffling (ablation)T5 · GLUEaverage score: 73.17 -> 82.96Improved
use BERT-style instead of prefix language modeling (ablation)T5 · GLUEaverage score: 80.69 -> 82.96Improved
replace MASS-style with random token swap (BERT-style) (ablation)T5 · GLUEaverage score: 82.32 -> 82.96No clear effect
drop corrupted tokens instead of replacing spans with sentinels (ablation)T5 · GLUEaverage score: 83.28 -> 84.44; SuperGLUE 71.36 -> 68.67Mixed
raise corruption rate from 15% to 50% (ablation)T5 · GLUEaverage score: 83.28 -> 81.27Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy